DESCRIPTION: (Applicant's Abstract) The Human Genome Project (HGP) has had astounding success in mapping. Unfortunately the progress in DNA sequencing has been slower. Although technologists have made great promises, no integrated systems are available to perform high-speed, high-throughput sequencing in an automated fashion. This proposal addresses the needs of the community for a modular sequencing system based upon two innovative but workable strategies. Template DNA will be prepared and sequencing reactions performed and purified using a solid phase reversible immobilization (SPRI) strategy. Gel-ready samples will be loaded from microtiter plates onto a capillary electrophoresis system in an automated fashion. To produce gel ready samples we will construct a fully automated system, based on our SPRI technology for DNA manipulation together with our existing devices and modules. We will use our SPRI approach to isolate the polymerase chain reaction (PCR) products, single or double stranded templates which will be the start-point for the sequencing process. SPRI will also be used for the concentration and desalting of samples prior to electrophoresis. Thermal cycle sequencing will be performed in our pre-existing devices which will be located on a robotics platform using an articulated gripper to move items between modules. The end- point will be samples ready to be loaded on to the Utah electrophoresis device. The availability of highly pure samples is especially important for a capillary electrophoresis (CE) instrument which is sensitive to the presence of salts, template DNA and other impurities for loading efficiency and matrix/coating stability. The SPRI technology will interface well with a CE based instrument. To perform electrophoresis, a capillary instrument will be built based upon Mathies' confocal laser induced fluorescence scanner. A four channel version of this scanner will be used to detect signal from 96 capillaries simultaneously. Loading of multiple capillaries has always been a difficult area to automate. In this proposal a system is described based upon fluid flow that will easily and reliably load samples onto 96 capillaries simultaneously. The loading system is inherently automated, simple to use and employs minimal robotics. The fluidic loading device is part of a complete fully automated 96 capillary sequencing instrument which is a modular solution to the large scale sequencing problem. Sequencing data will be interpreted using a neural network based algorithm that again is fully modular. The neural network software is expected to perform in real-time, interpreting sequence as the data is obtained. The individual system components will be fully tested and validated against other solutions to the same problems (Streptavidin-coated magnetic bead methods, ABI machines, multiplexing, other gel readers, etc.). The system as a whole will be capable of very high throughput with a minimum of operator intervention. It can satisfy the requirements of RFA HG-95-001 and the needs of the human genome community at large for improved reliable automated DNA sequencing technology.